All of lore.kernel.org
 help / color / mirror / Atom feed
From: Qu Wenruo <quwenruo@cn.fujitsu.com>
To: linux-btrfs@vger.kernel.org
Subject: [PATCH v2 05/19] btrfs-progs: Introduce wrapper to recover raid56 data
Date: Mon, 26 Dec 2016 14:29:25 +0800	[thread overview]
Message-ID: <20161226062939.5841-6-quwenruo@cn.fujitsu.com> (raw)
In-Reply-To: <20161226062939.5841-1-quwenruo@cn.fujitsu.com>

Introduce a wrapper to recover raid56 data.

The logical is the same with kernel one, but with different interfaces,
since kernel ones cares the performance while in btrfs we don't care
that much.

And the interface is more caller friendly inside btrfs-progs.

Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
---
 kernel-lib/raid56.c | 77 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 kernel-lib/raid56.h | 11 ++++++++
 2 files changed, 88 insertions(+)

diff --git a/kernel-lib/raid56.c b/kernel-lib/raid56.c
index e078972b..e3a9339e 100644
--- a/kernel-lib/raid56.c
+++ b/kernel-lib/raid56.c
@@ -280,3 +280,80 @@ int raid6_recov_datap(int nr_devs, size_t stripe_len, int dest1, void **data)
 	}
 	return 0;
 }
+
+/* Original raid56 recovery wrapper */
+int raid56_recov(int nr_devs, size_t stripe_len, u64 profile, int dest1,
+		 int dest2, void **data)
+{
+	int min_devs;
+	int ret;
+
+	if (profile & BTRFS_BLOCK_GROUP_RAID5)
+		min_devs = 2;
+	else if (profile & BTRFS_BLOCK_GROUP_RAID6)
+		min_devs = 3;
+	else
+		return -EINVAL;
+	if (nr_devs < min_devs)
+		return -EINVAL;
+
+	/* Nothing to recover */
+	if (dest1 == -1 && dest2 == -1)
+		return 0;
+
+	/* Reorder dest1/2, so only dest2 can be -1  */
+	if (dest1 == -1) {
+		dest1 = dest2;
+		dest2 = -1;
+	} else if (dest2 != -1 && dest1 != -1) {
+		/* Reorder dest1/2, ensure dest2 > dest1 */
+		if (dest1 > dest2) {
+			int tmp;
+
+			tmp = dest2;
+			dest2 = dest1;
+			dest1 = tmp;
+		}
+	}
+
+	if (profile & BTRFS_BLOCK_GROUP_RAID5) {
+		if (dest2 != -1)
+			return 1;
+		return raid5_gen_result(nr_devs, stripe_len, dest1, data);
+	}
+
+	/* RAID6 one dev corrupted case*/
+	if (dest2 == -1) {
+		/* Regenerate P/Q */
+		if (dest1 == nr_devs - 1 || dest1 == nr_devs - 2) {
+			raid6_gen_syndrome(nr_devs, stripe_len, data);
+			return 0;
+		}
+
+		/* Regerneate data from P */
+		return raid5_gen_result(nr_devs - 1, stripe_len, dest1, data);
+	}
+
+	/* P/Q bot corrupted */
+	if (dest1 == nr_devs - 2 && dest2 == nr_devs - 1) {
+		raid6_gen_syndrome(nr_devs, stripe_len, data);
+		return 0;
+	}
+
+	/* 2 Data corrupted */
+	if (dest2 < nr_devs - 2)
+		return raid6_recov_data2(nr_devs, stripe_len, dest1, dest2,
+					 data);
+	/* Data and P*/
+	if (dest2 == nr_devs - 1)
+		return raid6_recov_datap(nr_devs, stripe_len, dest1, data);
+
+	/*
+	 * Final case, Data and Q, recover data first then regenerate Q
+	 */
+	ret = raid5_gen_result(nr_devs - 1, stripe_len, dest1, data);
+	if (ret < 0)
+		return ret;
+	raid6_gen_syndrome(nr_devs, stripe_len, data);
+	return 0;
+}
diff --git a/kernel-lib/raid56.h b/kernel-lib/raid56.h
index e088279b..9aee39aa 100644
--- a/kernel-lib/raid56.h
+++ b/kernel-lib/raid56.h
@@ -44,4 +44,15 @@ int raid6_recov_data2(int nr_devs, size_t stripe_len, int dest1, int dest2,
 		      void **data);
 /* Recover data and P */
 int raid6_recov_datap(int nr_devs, size_t stripe_len, int dest1, void **data);
+
+/*
+ * Recover raid56 data
+ * @dest1/2 can be -1 to indicate correct data
+ *
+ * Return >0 for unrecoverable case.
+ * Return 0 for recoverable case, And recovered data will be stored into @data
+ * Return <0 for fatal error
+ */
+int raid56_recov(int nr_devs, size_t stripe_len, u64 profile, int dest1,
+		 int dest2, void **data);
 #endif
-- 
2.11.0




  parent reply	other threads:[~2016-12-26  6:30 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-12-26  6:29 [PATCH v2 00/19] Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 01/19] btrfs-progs: raid56: Introduce raid56 header for later recovery usage Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 02/19] btrfs-progs: raid56: Introduce tables for RAID6 recovery Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 03/19] btrfs-progs: raid56: Allow raid6 to recover 2 data stripes Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 04/19] btrfs-progs: raid56: Allow raid6 to recover data and p Qu Wenruo
2016-12-26  6:29 ` Qu Wenruo [this message]
2016-12-26  6:29 ` [PATCH v2 06/19] btrfs-progs: Introduce new btrfs_map_block function which returns more unified result Qu Wenruo
2017-02-24  0:37   ` Liu Bo
2017-02-24  0:45     ` Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 07/19] btrfs-progs: Allow __btrfs_map_block_v2 to remove unrelated stripes Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 08/19] btrfs-progs: csum: Introduce function to read out one data csum Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 09/19] btrfs-progs: scrub: Introduce structures to support fsck scrub for RAID56 Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 10/19] btrfs-progs: scrub: Introduce function to scrub mirror based tree block Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 11/19] btrfs-progs: scrub: Introduce function to scrub mirror based data blocks Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 12/19] btrfs-progs: scrub: Introduce function to scrub one extent Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 13/19] btrfs-progs: scrub: Introduce function to scrub one data stripe Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 14/19] btrfs-progs: scrub: Introduce function to verify parities Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 15/19] btrfs-progs: extent-tree: Introduce function to check if there is any extent in given range Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 16/19] btrfs-progs: scrub: Introduce function to recover data parity Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 17/19] btrfs-progs: scrub: Introduce a function to scrub one full stripe Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 18/19] btrfs-progs: scrub: Introduce function to check a whole block group Qu Wenruo
2016-12-26  6:29 ` [PATCH v2 19/19] btrfs-progs: fsck: Introduce offline scrub function Qu Wenruo
2016-12-26  8:42 ` [PATCH v2 00/19] Btrfs offline scrub Qu Wenruo
2016-12-29 18:15 ` [PATCH v2 00/19] Goffredo Baroncelli
2016-12-30  0:40   ` Qu Wenruo
2016-12-30 18:39     ` Goffredo Baroncelli
2017-01-03  0:25       ` Qu Wenruo

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161226062939.5841-6-quwenruo@cn.fujitsu.com \
    --to=quwenruo@cn.fujitsu.com \
    --cc=linux-btrfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.